@@ -86,6 +86,7 @@ import(grid)

 import(utils)

 importFrom(S4Vectors,as.data.frame)

 importFrom(S4Vectors,do.call)

+importFrom(S4Vectors,mcols)

 importFrom(dplyr,mutate)

 importFrom(dplyr,n)

 importFrom(rlang,sym)

@@ -6,7 +6,7 @@ pcaTrainInterface <- function(measurements, classes, params, nFeatures, ...)

               ###

               # Splitting measurements into a list of each of the datasets

               ###

-              assayTrain <- sapply(unique(mcols(measurements)[["assay"]]), function(assay) measurements[, mcols(measurements)[["assay"]] %in% assay], simplify = FALSE)

+              assayTrain <- sapply(unique(S4Vectors::mcols(measurements)[["assay"]]), function(assay) measurements[, S4Vectors::mcols(measurements)[["assay"]] %in% assay], simplify = FALSE)

               if(!"clinical" %in% names(assayTrain)) stop("Must have an assay called \"clinical\".")

@@ -45,7 +45,7 @@ prevalTrainInterface <- function(measurements, classes, params, ...)

               ###

               # Splitting measurements into a list of each of the assays

               ###

-              assayTrain <- sapply(unique(mcols(measurements)[["assay"]]), function(assay) measurements[, mcols(measurements)[["assay"]] %in% assay], simplify = FALSE)

+              assayTrain <- sapply(unique(S4Vectors::mcols(measurements)[["assay"]]), function(assay) measurements[, S4Vectors::mcols(measurements)[["assay"]] %in% assay], simplify = FALSE)

               if(!"clinical" %in% names(assayTrain)) stop("Must have an assay called \"clinical\"")

@@ -68,7 +68,7 @@

 #'             measurements[testIndices, ], classes[testIndices], modellingParams = modellingParams)

 #'   #}

 #' 

-#' @importFrom S4Vectors do.call

+#' @importFrom S4Vectors do.call mcols

 #' @usage NULL

 #' @export

 setGeneric("runTest", function(measurementsTrain, ...)

@@ -112,7 +112,7 @@ function(measurementsTrain, outcomeTrain, measurementsTest, outcomeTest,

     }

   }

-  if("feature" %in% colnames(mcols(measurementsTrain))) originalFeatures <- mcols(measurementsTrain)[, na.omit(match(c("assay", "feature"), colnames(mcols(measurementsTrain))))]    

+  if("feature" %in% colnames(S4Vectors::mcols(measurementsTrain))) originalFeatures <- S4Vectors::mcols(measurementsTrain)[, na.omit(match(c("assay", "feature"), colnames(S4Vectors::mcols(measurementsTrain))))]    

   else originalFeatures <- colnames(measurementsTrain)

   if(!is.null(modellingParams@selectParams) && max(modellingParams@selectParams@tuneParams[["nFeatures"]]) > ncol(measurementsTrain))

@@ -82,7 +82,7 @@ setMethod("runTests", "DataFrame", function(measurements, outcome, crossValParam

     stop("Some data elements are missing and classifiers don't work with missing data. Consider imputation or filtering.")            

   originalFeatures <- colnames(measurements)

-  if("feature" %in% colnames(mcols(measurements))) originalFeatures <- mcols(measurements)[, c("assay", "feature")]                 

+  if("feature" %in% colnames(S4Vectors::mcols(measurements))) originalFeatures <- S4Vectors::mcols(measurements)[, c("assay", "feature")]                 

   splitDataset <- prepareData(measurements, outcome)

   measurements <- splitDataset[["measurements"]]

   outcome <- splitDataset[["outcome"]]

@@ -107,7 +107,7 @@ input data. Autmomatically reducing to smaller number.")

   characteristics <- characteristics

   verbose <- verbose

   # Make them all local variables, so they are passed to workers.

-  

+

   results <- bpmapply(function(trainingSamples, testSamples, setNumber)

   #results <- mapply(function(trainingSamples, testSamples, setNumber)

   {

@@ -1,6 +1,6 @@

 selectMulti <- function(measurementsTrain, classesTrain, params, verbose = 0)

           {

-              assayTrain <- sapply(unique(mcols(measurementsTrain)[["assay"]]), function(assay) measurementsTrain[, mcols(measurementsTrain)[["assay"]] %in% assay], simplify = FALSE)

+              assayTrain <- sapply(unique(S4Vectors::mcols(measurementsTrain)[["assay"]]), function(assay) measurementsTrain[, S4Vectors::mcols(measurementsTrain)[["assay"]] %in% assay], simplify = FALSE)

               featuresIndices <- mapply(.doSelection, 

                                          measurements = assayTrain,

                                          modellingParams = params,